U.S. patent number 4,727,874 [Application Number 06/649,261] was granted by the patent office on 1988-03-01 for electrosurgical generator with high-frequency pulse width modulated feedback power control.
This patent grant is currently assigned to C. R. Bard, Inc.. Invention is credited to William J. Bowers, Phillip D. Hardwick.
United States Patent |
4,727,874 |
Bowers , et al. |
March 1, 1988 |
Electrosurgical generator with high-frequency pulse width modulated
feedback power control
Abstract
A pulse width modulation technique regulates the output power of
each cycle of a radio frequency surgical signal of an
electrosurgical generator. The delivered power of the surgical
signal is determined by multiplying the sensed current and the
sensed voltage of the surgical signal. An error signal is
established by the difference of the actual delivered power with
respect to a selected desired output power. The error signal is
operatively utilized to modulate the pulse width of each driving
pulse which creates the cycles of the surgical signal. Limits on
the sensed voltage and sensed current signals are established to
limit the output characteristics of the surgical signal. A minimum
current limit signal is utilized to limit the maximum output
voltage into relatively high impedances. A minimum voltage limit
signal is utilized to limit the maximum output current into
relatively low impedances. Very rapid response times and very
effective power regulation even into relatively high impedance
tissues are possible with the pulse width modulation technique. The
risks and problems associated with open circuit flashing, alternate
path burns and closed circuit shorting are substantially reduced or
eliminated.
Inventors: |
Bowers; William J. (Aurora,
CO), Hardwick; Phillip D. (Aurora, CO) |
Assignee: |
C. R. Bard, Inc. (Murray Hill,
NJ)
|
Family
ID: |
24604077 |
Appl.
No.: |
06/649,261 |
Filed: |
September 10, 1984 |
Current U.S.
Class: |
606/38; 330/251;
606/39; 330/207A; 606/37; 606/40 |
Current CPC
Class: |
A61B
18/1206 (20130101); H03F 1/52 (20130101); H02M
7/53871 (20130101); A61B 2018/00779 (20130101); A61B
2018/00827 (20130101); A61B 2018/00892 (20130101) |
Current International
Class: |
A61B
18/12 (20060101); H02M 7/5387 (20060101); A61B
017/39 () |
Field of
Search: |
;128/303.12,303.13,303.14,303.15,303.17,303.18 ;330/251,27A |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
0136855 |
|
Apr 1986 |
|
EP |
|
1519225 |
|
Jun 1978 |
|
GB |
|
2085243 |
|
Apr 1982 |
|
GB |
|
2108786 |
|
May 1983 |
|
GB |
|
Primary Examiner: Cohen; Lee S.
Assistant Examiner: Shay; David
Attorney, Agent or Firm: Ley; John R.
Claims
What is claimed is:
1. In a electrosurgical generator including means for supplying a
surgical signal at a predetermined high frequency to perform a
surgical procedure, and means for regulating the power content of
the surtgical signal, and an improved feedback means for
controlling the power regulating means comprising:
means for creating a current delivered signal related to the
current of the surgical signal;
means for establishing a current limit signal;
voltage limit means receptive of the current limit signal and the
current delivered signal and operative for supplying a first signal
corresponding to one of either the current limit signal or the
current delivered signal, said voltage limit means operatively
supplying the current delivered signal as the first signal when the
current delivered signal is greater than the current limit signal
and operatively supplying the current limit signal as the first
signal when the current delivered signal is less than the current
limit signal;
means for creating a voltage delivered siganl related to the
voltage of the surgical signal;
means for establishing a voltage limit signal;
curret limit means receptive of the voltage limit signal and
voltage delivered signal and operative for supplying a second
signal corresponding to one of either the voltage limit signal or
the voltage delivered signal, said current limit means operatively
supplying the voltage delivered signal as the second signal when
the voltage delivered signal is greater than the voltage limit
signal and operatively supplying the voltage limit signal as the
second signal when the voltage delivered signal is less than the
voltage limit signal;
multiplier means receptive of the first and second signals and
operative for multiplying the first and second signals to create a
delivered power signal which is the product of the first and second
signals; and
means receptive of the delivered power signal and operative for
controlling the power regulation means of said generator to
regulate the power content of the surgical signal in response to
the power delivered signal.
2. In an electrosurgical generator as defined in claim 1, the
current delivered signal is directly related to the RMS current of
the surgical signal, and the voltage delivered signal is directly
related to the RMS voltage of the surgical signal.
3. In an electrosurgical generator as defined in claim 1, at least
one of the limit signals received by said limit means is of a
constant value.
4. In an electrosurgical generator as defined in claim 3, the
current delivered signal is directly related to the RMS current of
the surgial signal, and the voltage delivered signal is directly
related to the RMS voltage of the surgical signal.
5. In an electrosurgical generator as defined in claim 1, said
improved feedback means further comprises:
means for supplying a desired output power signal representative of
a predetermined amount of power which the surgical signal is
desired to contain;
differential means receptive of the delivered power signal and the
desired output power signal and operative for creating an error
signal representative of the difference of the delivered power
signal with respect to the desired output power signal; and
wherein:
said means for controlling the power regulation means does so in
predetermined relation to the error signal.
6. In an electrosurgical generator as defined in claim 5, at least
one of the limit signals received by said limit means is related to
the desired output power signal.
7. In an electrosurgical generator as defined in claim 6, the
current delivered signal is directly related to the RMS current of
the surgical signal, and the voltage delivered signal is directly
related to the RMS voltage of the surgical signal.
8. In an electrosurgical generator as definned in claim 6, the one
limit signal received by said limit means is non-linearally related
to the desired output power signal.
9. In an electrosurgical generator as defined in claim 6, the
current delivered signal is directly related to the RMS current of
the surgical signal, and the voltage delivered signal is directly
related to the RMS voltage of the surgical signal.
10. In an electrosurgical generator as defined in claim 1, said
surgical signal is a series of substantially sinusoidally shaped
cycles occurring at the predetermined high frequency, and a further
improvement in said means for supplying the surgical signal
comprises:
drive means for creating a periodic series of driving pulses
occurring at a predetermined frequency which is related to the
predetermined high frequency of the surgical signal, each driving
pulse having a predetermined energy content related to the time
width of the driving pulse; and
surgical signal creating means receptive of the driving pulses for
creating the surgical signal from the driving pulses, said creating
means creating each cycle of the surgical signal from at least one
corresponding driving pulse, said creating means further
establishing an energy content of each cycle of the surgical signal
in a direct relationship to the energy content of each
corresponding driving pulse from which the cycle of the surgical
signal was created.
11. In an electrosurgical generator as defined in claim 10, the
predetermined frequency at which the driving pulses occur is twice
the predetermined high frequency of the surgical signal, alternate
driving pulses in the periodic series primarily create a positive
half cycle of each cycle of the surgical signal and the other
alternate driving pulses primarily create the negative half cycle
of each cycle of the surgical signal, and sequential driving pulses
alternate in polarity with respect to one another.
12. In an electrosurgical generator as defined in claim 11, said
means for supplying the surgical signal further comprises:
bandpass filter means receptive of the driving pulses and operative
for converting the driving pulses into the substantially
sinusoidally shaped cycles of the surgical signal at the
predetermined high frequency, said bandpass filter means
operatively changes the amplitude of the substantially sinusoidally
shaped cycles of the surgical signal in a predetermined
relationship to the width of each driving pulse creating the cycle
of the surgical signal.
13. An electrosurgical generator which supplies a surgical signal
having a series of substantially sinusoidually shaped cycles
occurring at a fixed predetermined radio frequency, comprising:
drive means for creating a periodic series of driving pulses
occurring at a predetermined frequency related to the predetermined
radio frequency of the surgical signal, each driving pulse having a
predetermined energy content related to the time width of the
driving pulse;
surgical signal creating means receptive of the driving pulses for
creating the surgical signal from the driving pulses, said creating
means creating each cycle of the surgical signal from at least one
corresponding driving pulse, said creating means further
establishing an energy content of each cycle of the surgical signal
in a direct relationship to the energy content of each
corresponding driving pulse from which the cycle of the surgical
signal was created;
means responsive to the surgical signal and operative for creating
a delivered power signal related to the power content of the
surgical signal;
means for establishing a desired output power signal related to a
desired amount of output power for the surgical signal; and
means receptive of the delivered power signal and the desired
output power signal and operative for modulating the width of the
driving pulses in accordance with a predetermined relationship of
the delivered power signal and the desired output power signal to
regulate the power content of the surgical signal.
14. An electrosurgical generator as defined in claim 13, wherein
the energy content of each cycle of the surgical signal is
established by varying the amplitude of the sinusoidal shaped
cycle.
15. An electrosurgical generator as defined in claim 13 further
comprising:
bandpass filter means receptive of the driving pulses and operative
for converting the driving pulses into the substantially
sinusoidally shaped cycles of the surgical signal at the
predetermined radio frequency.
16. An electrosurgical generator as defined in claim 15
wherein:
said bandpass filter means operatively changes the amplitude of the
substantially sinusoidally shaped cycles of the surgical signal in
a predetermined relationship to the width of each driving pulse
creating the cycle of the surgical signal.
17. An electrosurgical generator as defined in claim 16
wherein:
said bandpass filter means primarily creates each half cycle of
each substantially sinusoidally shaped cycle of the surgical signal
from a corresponding drivig pulse of the periodic series of driving
pulses.
18. An electrosurgical generator as defined in claim 16
wherein:
the predetermined frequency at which the driving pulses occur is
twice the predetermined radio frequency of the surgical signal,
alternate driving pulses of the periodic series primarily create a
positive half cycle of each cycle of the surgical signal and the
other alternate driving pulses of the periodic series create the
negative half cycle of each cycle of the surgical signal, and
sequential driving pulses of the periodic series alternate in
polarity with respect to one another.
19. An electrosurgical generator as defined in claim 13 further
comprising:
means for creating one of a current sensed signal or a voltage
sensed signal related to the current or the voltage content of the
surgical signal supplied by the electrosurgical generator,
respectively;
means for establishing one of a current limit signal or a voltage
limit signal;
limit means receptive of the one limit signal and the one sensed
signal which have the same current or voltage relationship and
operative for supplying the sensed signal as a delivered signal
when the one sensed signal occupies a first predetermined
relationship to the one limit signal and operative for supplying
the limit signal as the delivered signal when the one sensed signal
occupies a second predetermined different relationship to the one
limit signal; and
and wherein said modulating means modulates the width of the
driving pulses in relation to the delivered signal.
20. An electrosurgical generator as defined in claim 19 wherein the
one limit signal established is of a constant value.
21. An electrosurgical generator as defined in claim 19 wherein the
one limit signal established is non-linearly related to a desired
amount of output power to which the surgical signal is to be
regulated.
22. An electrosurgical generator as defined in claim 19 wherein the
one limit signal established is linearly related to a desired
amount of output power to which the surgical signal is to be
regulated.
23. An electrosurgical generator as defined in claim 13 wherein the
delivered power signal is related to the instantaneous RMS power
content of the surgical signal.
24. An electrosurgical generator which supplies a predetermined
surgical signal to perform a surgical procedure and which regulates
the power content of the surgical signal, the surgical signal being
a series of individual cycles occurring at a predetermined radio
frequency, said generator comprising:
drive means for creating a drive signal defined by a periodic
series of driving pulses occurring at a predetermined frequency and
time relationship with respect to each cycle of the radio frequency
surgical signal, each driving pulse having a predetermined energy
content related to the time width of the driving pulse;
surgical signal creating means receptive of the driving pulses for
creating the surgical signal from the driving pulses, said creating
means creating each cycle of the surgical signal from at least one
corresponding driving pulse, said creating means further
establishing an energy content of each cycle of the surgical signal
in a direct relationship to the energy content of each
corresponding driving pulse from which the cycle of the surgical
signal was created;
means for creating a current delivered signal related to the RMS
current of the surgical signal;
means for establishing a current limit signal;
voltage limit means receptive of the current limit signal and the
current delivered signal and operative for supplying a first signal
corresponding to one of either the current limit signal or the
current delivered signal, said voltage limit means operatively
supplying the current delivered signal as the first signal when the
current delivered signal is greater than the current limit signal
and operatively supplying the current limit signal as the first
signal when the current delivered signal is less than the current
limit signal;
means for creating a voltage delivered signal related to the RMS
voltage of the surgical signal;
means for establishing a voltage limit signal;
current limit means receptive of the voltage limit signal and
voltage delivered signal and operative for supplying a second
signal corresponding to one of either the voltage limit signal or
the voltage delivered signal, said current limit means operatively
supplying the voltage delivered signal as the second signal when
the voltage delivered signal is greater than the voltage limit
signal and operatively supplying the voltage limit signal as the
second signal when the voltage delivered signal is less than the
voltage limit signal;
multiplier means receptive of the first and second signals and
operative for multiplying the first and second signals to create a
delivered power signal which is the product of the first and second
signals;
means for supplying a desired output power signal representative of
a predetermined amount of power which the surgical signal is
desired to contain;
differential means receptive of the delivered power signal and the
desired output power signal and operative for creating an error
signal representative of the difference of the delivered power
signal with respect to the desired output power signal; and
modulation means receptive of the error signal and operative for
controlling said drive means to modulate the width of each driving
pulse in a predetermined relation to the error signal, said
modulation means operatively changing the predetermined energy
content of each driving pulse to regulate the energy content of
each cycle of the surgical signal to a level related to the power
level represented by the desired output power signal.
25. A electrosurgical generator as defined in claim 24 wherein said
modulation means further comprises:
integrator means receptive of the error signal and operative for
integrating the error signal over time and creating a trigger level
signal related to the integrated value of the error signal;
means for creating a ramp signal having a periodic series of ramp
waveforms occurring at a predetermined frequency related to the
frequency of the driving pulses;
comparator means receptive of the ramp signal and the trigger level
signal and operative for creating a pulse width control signal
having a characteristic occurring periodically at the predetermined
frequency of the ramp signal, the pulse width control signal
operatively controlling the width of each driving pulse.
26. An electrosurgical generator as defined in claim 25 wherein
said drive means further comprises:
pulse phase means for creating a pulse phase signal having a
periodic series of phase pulses occurring at the predetermined
frequency of said driving pulses; and
gating means receptive of the pulse phase signal and the pulse
width control signal and operative for creating each driving pulse
having a time width related to the phase pulse signal and the
periodic characteristic of the pulse width control signal.
27. An electrosurgical generator as defined in claim 26 wherein
each phase pulse signal has a predetermined time width and the
width of each phase pulse signal defines the maximum possible width
of each driving pulse.
28. An electrosurgical generator as defined in claim 27
wherein:
said gating means operatively initiates each driving pulse in
relation to the occurrence of each phase pulse and operatively
terminates each driving pulse in relation to the occurrence of the
periodic characteristic of the pulse width control signal.
29. An electrosurgical generator as defined in claim 26
wherein:
said pulse phase means creates a pulse phase one signal and a pulse
phase two signal which are phase shifted with respect to one
another by one hundred eighty degrees, both the pulse phase one
signal and the pulse phase two signal having the characteristics of
the aforesaid pulse phase signal;
the predetermined frequency of the ramp waveforms of the ramp
signal and of the periodic characteristic of the pulse width
control signal are two times the frequency of the surgical signal;
and
said gating means is receptive of the pulse phase one signal and
the pulse phase two signal and operatively creates individual phase
one driving pulses in relation to the phase one pulse signal and
the periodic characteristic of the pulse width control signal and
operatively creates individual phase two driving pulses in relation
to the phase two pulse signal and the periodic characteristic of
the pulse width control signal, each phase one driving pulse and
each phase two driving pulse having the characteristics of each
aforesaid driving pulse, the phase one driving pulses and the phase
two driving pulses defining the drive signal.
30. An electrosurgical generator as defined in claim 29
wherein:
said surgical signal creating means receptive of the driving pulses
and operative for creating each cycle of the surgical signal
operatively creates one half-cycle of each cycle of the surgical
signal from a phase one driving pulse and operatively creates the
other half-cycle of each cycle of the surgical signal from a phase
two driving pulse.
31. An electrosurgical generator as defined in claim 24 wherein
each cycle of the surgical signal is substantially sinusoidally
shaped and said surgical signal creating means further
comprises:
bandpass filter means receptive of the driving pulses and operative
for converting the driving pulses into the substantially
sinusoidally shaped cycles of the surgical signal at the
predetermined radio frequency.
32. An electrosurgical generator as defined in claim 31
wherein:
said bandpass filter means substantially inhibits frequency
components of the driving pulses at other than the predetermined
radio frequency.
33. An electrosurgical generator as defined in claim 31
wherein:
said bandpass filter means operatively changes the amplitude of the
substantially sinusoidally shaped cycles of the surgical signal in
a predetermined relationship to the width of each driving pulse
creating the cycle of the surgical signal.
34. An electrosurgical generator as defined in claim 31
wherein:
said bandpass filter means primarily creates each half cycle of
each substantially sinusoidally shaped cycle of the surgical signal
from one corresponding driving pulse.
Description
This invention pertains to an electrosurgical generator having an
improved output power regulation capability as a result of a closed
loop feedback power control network utilizing pulse width
modulation at the frequency of, and to control the energy content
of, each cycle of the high-frequency surgical signal, among other
improved features.
By use of an electrosurgical generator in a surgical procedure, it
is possible for the surgeon to cut, to blend or cut with
hemostasis, or to purely coagulate. The surgeon can quickly select
and change the different modes of operation as the surgical
procedure progresses. In each mode of operation, it is important to
regulate the electrical power delivered to the patient to achieve
the desired surgical effect. Applying more power than is necessary
will result in unnecessary tissue destruction and prolong healing.
Applying less than the desired amount of electrical power will
usually inhibit the surgical procedure. Different types of tissues
will be encountered as the procedure progresses and each different
tissue will usually require more or less power due to a change in
inherent tissue impedance. Accordingly, all successful types of
electrosurgical generators employ some type of power regulation in
order to control the electrosurgical effects desired by the
surgeon.
Two types of power regulation are conventional in previous
electrosurgical generators. The most common type controls the DC
power supply of the generator. This type of power regulation limits
the amount of power absorbed from the conventional AC power line to
which the generator is connected. A feedback control loop compares
the actual power supplied by the power supply to a desired power
setting in order to achieve regulation. Another type of power
regulation in previous electrosurgical generators involves
controlling the gain of the high-frequency or radio frequency
amplifier. A feedback control loop compares the output power
supplied from the RF amplifier to a desired power level, and the
gain is adjusted accordingly.
While both known types of power regulation have achieved moderate
success, there nevertheless have been certain undesirable
characteristics associated with each. One undesirable
characteristic involves the response time for regulation. The
impedance of the different tissues encountered during the surgical
procedure can fluctuate substantially. In moving from a high
impedance tissue to a low impedance tissue, the low impedance
tissue may be unnecessarily destroyed or damaged before the
electrosurgical generator can reduce the output power to a level
compatible with the low impedance tissue. Similarly, when a high
impedance tissue is encountered, the output power from the
generator may be momentarily insufficient to create or continue the
precise surgical effect desired by the surgeon. Precise execution
of the surgical procedure becomes difficult or impossible.
Another problem of power regulation in previous electrosurgical
generators has resulted in large measure because such previous
generators have been designed to attain maximum power transfer to
intermediate impedance ranges. As with any amplifier, an
electrosurgical generator will achieve maximum power transfer when
its internal impedance is equal to the output load impedance to
which the generator is connected. At high impedances, the power
delivered inherently rolls off because of the difference in load
impedance compared to the internal impedance. To compensate, the
surgeon increases the power setting to a higher level than
necessary. As soon as the incision progresses through the high
impedance tissue, the output power is too great and tissue
destruction or undesirable surgical effects result. Making the
initial incision is an example. The skin includes a relatively
large percentage of dead cells and cells which contain considerably
less moisture than other cells in tissues beneath the skin, which
increases its impedance compared to the impedance of the tissues
below the skin. A higher power setting is therefore required for
the initial incision. However, as soon as the incision is made, a
reduced amount of power is all that is necessary. With typical
previous electrosurgical generators, the initial incision was
deeper than desired because the active electrode, i.e., the
electrosurgical instrument, went deeper than the surgeon desired
due to the excessive amount of power delivered. The surgeon usually
desires to control the depth of the incision and conduct the
surgical procedure in controlled depth levels. If the power
regulation is not reliable, a deeper incision in certain areas may
result in undesired bleeding or other undesirable surgical effects.
It is for this reason that most surgeons generally prefer to make
the initial incision using a conventional scalpel, rather than
using the active electrode of an electrosurgical generator.
Another power-regulation-related problem of previous
electrosurgical generators is open circuit flashing just prior to
the commencement of the surgical procedure. Before the
electrosurgical procedure commences, no output power is supplied
due to the open circuit condition. The regulation circuitry
attempts to compensate by creating maximum power delivery
situation. As soon as the active electrode is moved into operative
distance from the tissue, an immediate flash or arcing is caused by
the relatively high voltage which exists due to the maximum power
delivery capability created by the power regulation circuitry.
Although continual arcing is desired in the coagulation
(fulguration) mode of operation, it is usually undesirable in the
other modes of operation. The power regulation circuitry eventually
compensates for the excessive power and reduces it. Nonetheless,
the initial arcing or flash usually causes excessive tissue
destruction and other undesirable tissue effects. The flash and
excessive tissue destruction can occur anytime the surgeon moves
the active electrode to the tissue.
Open circuit or excessively high output impedance conditions also
increase the risks of alternate path burns to the patient.
Alternate path burns are burns created by current flowing from the
patient to some surrounding grounded conductive object such as the
surgical table, rather than returning to the electrosurgical
generator through the patient plate, i.e., the inactive electrode.
Alternate path burns usually are caused by radio frequency leakage
currents created by the high-frequency surgical signal flowing
through stray capacitances between the patient and an adjacent
grounded object. Reducing the output voltage under open circuit or
high impedance conditions reduces the magnitude of and possibility
for radio frequency leakage currents.
Another power-regulation-related problem of previous
electrosurgical generators relates to shorting the output terminals
of the generator. Human nature being what it is, one usual,
although not recommended, technique of quickly determining whether
an electrosurgical generator is operating is to simply short the
two output electrodes and observe an electrical spark. A not
unusual result of such shorting is the destruction of the power
supply in the generator. The generator is forced to quickly attempt
to regulate from a high power open circuit condition to a short
circuit low impedance condition. Due to the limitations on
regulating capability, the electrical power components of the power
supply are usually overdriven and are quickly destroyed before
compensation can occur.
BRIEF SUMMARY OF THE INVENTION
The present invention teaches an improved technique of regulating
the output power of an electrosurgical generator which obtains a
more rapid response time to obtain better and constant power
regulation even into relatively high and low impedance loads, and
which limits the output current and voltage to avoid or reduce the
problems of and risks associated with open circuit flashing,
alternate path burns and short circuit destructive currents.
In accordance with one of its major aspects, each cycle of a
high-frequency surgical signal supplied by the electrosurgical
generator is regulated in power content by modulating the width of
driving pulses of energy. The driving pulses operatively create
each cycle of the surgical signal. A closed loop feedback power
control arrangement creates a delivered power signal representative
of the power content of the surgical signal by sensing the current
and voltage associated with the surgical signal. The width of each
driving pulse of energy is modulated in accordance with a
relationship of the delivered power signal relative to a selected
desired output power signal to thereby regulate the power content
of the surgical signal to an amount substantially equivalent to the
desired amount of output power. Since each cycle of the surgical
signal is regulated in power content, very rapid power regulation
response times are possible. At desired output power levels which
are less than the full capacity of the electrosurgical generator,
power regulation and control is attained even into relatively high
impedance tissues, unlike previous electrosurgical situations where
power roll-off and lack of regulation typically occurred.
In accordance with another improved aspect, a voltage or a current
limit signal is substituted for the actual sensed voltage or
current signal in order to, respectively, limit the maximum output
current of the generator into relatively low impedances and limit
the maximum output voltage of the generator into relatively high
impedances. Limiting the maximum output voltage into relatively
high impedances attains the desirable effects of reducing or
eliminating flash and undesirable arcing, of achieving beneficial
electrosurgical effects on the tissue, and of reducing the risk of
alternate path burns. Limiting the maximum output current into
relatively low impedances has the beneficial effect of preventing
destructively high currents, even when short circuiting of the
output terminals or electrosurgical electrodes of the
generator.
The actual aspects of the present invention are defined in the
appended claims. A more complete understanding of the improvements
of the electrosurgical generator can be obtained from the following
description of a preferred embodiment taken in conjunction with the
drawings.
DETAILED DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram of the electrosurgical generator of the
present invention.
FIG. 2 is an expanded block and schematic diagram of certain
portions of FIG. 1.
FIG. 3 is an expanded block and schematic diagram of certain
portions of FIG. 1.
FIGS. 4A, 4B, 4C, 4D, 4E, 4F, 4G, 4H, 4I, 4J, 4K, 4L and 4M are
waveform diagrams illustrating signals present at certain locations
in the diagrams shown in FIGS. 1 and 3.
FIG. 5 is a graph of output surgical signal power relative to
output (tissue) impedance illustrating power regulation curves
attained by the circuit arrangement illustrated in FIG. 2.
FIG. 6 is a graph of output surgical signal power relative to
output (tissue) impedance of the electrosurgical generator when
modifications to a portion of the circuit shown in FIG. 2 are
made.
FIG. 7 is a schematic diagram of a circuit intended to replace a
portion of a circuit shown in FIG. 2.
DESCRIPTION OF THE PREFERRED EMBODIMENT
A preferred embodiment of the electrosurgical generator of the
present invention is shown and referenced 10 in FIG. 1. A control
panel 12 of the generator 10 includes the typical switches and
other control devices for controlling the mode of operation of the
generator 10 and the amount of power to be delivered in each mode.
In addition, the control panel 12 may include means for adjusting
the blend or relative amounts of cutting and hemostasis which
occurs during the cutting with hemostasis mode of operation. AC
power is supplied to the generator 10 from a conventional AC power
line 14. A controllable DC power supply 16 converts the AC power
from the line 14 to a DC power level at 20. A power output control
signal is supplied at 18 from the control panel 12 to control and
generally limit the DC power output at 20 from the supply 16
according to the amount of power desired. The output power at 20
from the supply 16 is applied to a conventional high frequency or
radio frequency electrosurgical amplifier 22. The amplifier 22
converts the DC power at 20 into a periodic pulse width modulated
signal at 24. A power transformer 26 receives the pulse width
modulated signal at 24 and converts it to an alternating or AC
pulse width modulated signal at 28.
The alternating pulse width modulated signal at 28 is applied to a
band pass filter 30 which has a band pass characteristic only at
the predetermined high or radio frequency of the surgical signal
delivered by the generator. The filter supplies a high frequency
surgical signal at 32. The surgical signal 32 creates the surgical
effect or procedure. The frequency of the surgical signal is
sufficiently high to avoid stimulating nerves, for example five
hundred kilohertz. The filter 30 eliminates any higher order
harmonics created by the amplifier 22 or the transformer 26 to
reduce the risk of alternate path (leakage capacitance) burns to
the patient. The filter 30 also inhibits the existence of
circulating DC currents created by rectification effects of the
tissue. The filter 30 converts the alternating signal at 28 to a
sinusoidal waveform due to the effects from the passive reactive
elements of the filter. The high-frequency surgical signal is
applied to conductor 32, which is connected to the active electrode
used by the surgeon. Conductor 34 is the reference potential
conductor for the high-frequency surgical signal and it is
connected to the patient plate or inactive electrode upon which the
patient is positioned. When a bipolar electrosurgical instrument is
used, both conductors 32 and 34 are connected to the instrument.
Although not shown, output isolation capacitors can be placed in
conductors 32 and 34 to also inhibit the DC circulating
currents.
A current sensor 36 is connected in series in the conductor 32 for
the purpose of deriving an instantaneous current sense signal at 38
which is related to the instantaneous magnitude of current flowing
in the conductor 32. A voltage sensor 40 is electrically connected
between the conductors 32 and 34 for the purpose of deriving an
instantaneous voltage sense signal at 42 representative of the
instantaneous voltage existing between the conductors 32 and 34.
Accordingly, both the instantaneous output current and voltage of
the high-frequency surgical signal are sensed at a point in the
generator 10 where the surgical signal is delivered. An accurate
indication of the amount of instantaneous output current and
voltage applied to the the tissue is thereby obtained. More exact
sense signals are obtained as compared to some prior arrangements
of sensing either current or voltage or both as they are applied to
the input terminal to an amplifier or to the input terminal of an
output transformer or the like. These prior arrangements suffer the
substantial disadvantages of failing to consider losses and
inefficiencies inherent in elements such as the amplifiers and
transformers.
To achieve individual pulse and cycle energy regulation of the
high-frequency surgical signal, the current and voltage sense
signals at 38 and 42 are applied to RMS to DC converters 44 and 46,
respectively. The converters 44 and 46 convert the input sense
signals to an RMS value represented by a DC output signal.
Accordingly, the signal present at 48 is a DC signal which
represents the RMS value of the actual output current of the
surgical signal, and the signal present at 50 is a DC signal which
represents the RMS value of the actual output voltage of the
surgical signal applied to the patient. Converting the
instantaneous current and voltage sense signals to RMS value gives
a true and accurate representation of the amount of current and
voltage actually delivered in the surgical signal, unlike other
prior techniques not involving RMS conversion.
The RMS current-related signal at 48 is applied to a current limit
circuit 52, and the RMS voltage-related signal at 50 is applied to
a voltage limit circuit 54. Minimum current limit and minimum
voltage limit signals at 56 and 58 are supplied to the limit
circuits 52 and 54, respectively, from a scaling circuit 60. The
scaling circuit 60 is operatively controlled by a mode logic
circuit 62 which supplies scaling control signals at 64 to the
scaling circuit 60. The scaling circuit 60 is also operatively
controlled by a selected power signal 66 supplied by the control
panel 12. The mode logic circuit 62 is controlled by mode control
signals applied at 65 from the control panel 12. The mode control
signals at 65 operatively establish the mode of operation of the
generator 10. The mode logic circuit 62 also supplies a control
signal at 67 to the power supply 16 to control the level of DC
power at 20 in accordance with the mode of operation selected.
The magnitude of the minimum current limit signal at 56 and the
magnitude of the minimum voltage limit signal at 58 are established
by the mode of operation of the generator 10, and in response to
the magnitude of the selected power signal applied at 66. The
minimum current limit signal at 56 represents a minimum amount of
current which is considered to be delivered into high impedances,
and has the effect of limiting the maximum voltage of the surgical
signal applied to high impedances. The minimum voltage limit signal
at 58 represents that magnitude of output voltage which is
considered to be delivered into low impedances, and has the effect
of limiting the maximum current of the surgical signal into low
impedances.
The limit circuit 52 compares the minimum current limit signal at
56 with the signal at 48 representative of the actual amount of
current delivered in the surgical signal. So long as the RMS
current-related signal at 48 exceeds the minimum current limit
signal at 56, the current limit circuit 52 supplies a current
delivered signal at 68 which corresponds to the signal at 48.
Similarly, the voltage limit circuit 54 compares the minimum
voltage limit signal at 58 with the signal at 50 representative of
the actual delivered voltage of the high-frequency surgical signal.
So long as the RMS voltage-related signal at 50 exceeds the minimum
voltage limit signal at 58, a voltage delivered signal is present
at 70 which corresponds to the signal at 50. Should either the RMS
current-related signal or the RMS voltage-related signal fall below
the levels of the signals at conductors 56 and 58, respectively,
the minimum current limit signal or the minimum voltage limit
signal is clamped and supplied at 68 or 70, respectively, as the
current delivered signal or the voltage delivered signal.
Accordingly, the current delivered signal at 68 is the greater one
of either the RMS current-related signal at 48 or the minimum
current limit signal present at 56. Similarly, the voltage
delivered signal at 70 is the greater of either the RMS
voltage-related signal at 50 or the minimum voltage limit signal at
58. Limiting the current delivered signal at 68 to a value no less
than that signal at 56 has the effect of holding the output voltage
of the surgical signal to a predetermined maximum level at high
impedances. Limiting the voltage delivered signal at 70 to the
minimum amount established by the signal at 58 has the effect of
limiting the output current of the surgical signal at low
impedances to a preestablished and safe maximum.
A signal representative of the delivered power is created by a
conventional analog multiplier 72, by multiplying the current
delivered signal at 68 and the voltage delivered signal at 70. The
multiplier 62 supplies a delivered power signal at 74.
The scaling circuit 60 also supplies a signal at 76 representative
of a desired output power level of the surgical signal. The scaling
circuit 60 establishes the desired output power signal at 76 in
accordance with the selected power signal at 66 from the control
panel 12, and in accordance with scaling control signals at 64
supplied by the mode logic circuit 62 according to the selected
mode of operation.
The desired output power signal at 76 and the delivered power
signal at 74 are compared at a differential amplifier 78 and an
error signal is supplied at 80. The error signal at 80 represents
the difference in magnitude between the delivered power and the
desired power. A pulse width modulation circuit 82 receives the
error signal at 80 and utilizes the error signal to create a pulse
width control signal at 84.
An amplifier drive circuit 86 receives the pulse width control
signal at 84 and creates a drive signal at 90. The drive signal is
defined by a series of driving pulses delivered at a predetermined
frequency to establish the predetermined frequency of the surgical
signal. The width or time duration of each driving pulse is
controlled by the pulse width control signal at 84. The drive
signal at 90 controls the operation of the amplifier 22. Each
driving pulse establishes the width and hence energy content of
each pulse of the pulse width modulated signal at 24. The width of
each pulse of the pulse width modulated signal regulates the output
power of each cycle of the surgical signal. Thus, this power is
ultimately controlled by the pulse width control signal at 84.
A duty cycle generator 92 is controlled by a signal at 94 from the
mode logic circuit 62. A duty cycle signal at 96 from the duty
cycle generator 94 also controls the amplifier drive circuit 86. A
duty cycle type of operation is typically established in the cut
with hemostasis and the coagulation modes of operation of the
generator 10. The duty cycle signal at 96 causes the amplifier
drive circuit 86 to control the delivery of pulses in the driving
signal at 90 in a periodic duty cycle fashion in accordance with
the mode of operation. In the cut mode of operation, the surgical
signal is a continuous sinusoidal wave and the duty cycle generator
92 is inoperative. A synchronization or oscillator signal is
supplied at 98 by the amplifier drive circuit 86 to cause the pulse
width modulation circuit 82 to synchronously respond at the same
frequency as the frequency of the driving pulses of the drive
signal at 90.
It is appreciated, therefore, that the pulse width control signal
at 84 is derived by a comparison of the delivered power signal to
the selected desired output power signal. Minor fluctuations in the
output level at 20 of the controllable DC power supply 16 become
largely insignificant because the primary or refined power control
is obtained by pulse width modulation. The number of components of
the main DC power supply can be reduced, as well as the cost of the
power supply and the size and weight of the electrosurgical
generator. For example, the typical expensive, heavy and costly
line transformer of the typical previous electrosurgical generator
power supply can be essentially eliminated and replaced by
controllable phase angle switching devices for controlling the
amount of power conducted from the AC line directly to the typical
rectifiers and filter capacitors. A power supply of reduced cost,
components, size and weight results, but it is still operatively
sufficient to obtain a sufficient amount of coarse power regulation
at the power supply 16 to allow the pulse width modulation
technique to achieve final precise power regulation.
Because of each of the driving pulses at the predetermined high
frequency is width and energy modulated, the power regulation
response times are rapid. The surgeon can more accurately and
precisely control the surgical procedure as it progresses, and many
of the previous typically-occurring undesirable effects caused by
tissue impedance changes can be substantially reduced or
eliminated.
Details of the RMS to DC converters 44 and 46, the limit circuits
52 and 54, the multiplier 72, the comparator 78, and the scaling
circuit 60 are shown in FIG. 2.
The selected power signal at 66 is derived by adjustment of a
conventional potentiometer (not shown) at the control panel 12
(FIG. 1). The selected power signal at 66 is a voltage signal which
represents the desired level of power. The selected power signal is
utilized to create the minimum current limit signal at 56 which is
applied to the limit circuit 52. The minimum current limit signal
at 56 is created by applying the selected power signal at 66 to an
operational amplifier (op amp) referenced 100. A conventional
square root network 102 is connected between the output terminal of
the op amp 100 and its input terminal which receives the selected
power signal at 66. The output signal from the op amp 100, present
at 104, generally represents the square root of the selected power
signal at 66. The square root of the selected power signal is
desired because the minimum current limit signal at 56 operatively
acts to control and limit the output voltage of the surgical signal
to a maximum constant level into high impedances. The output
voltage of the surgical signal is related to the output power by a
squared function for a given impedance or resistance load and thus
the output power is related to the output voltage by a square root
function. Accordingl.y, since the selected power signal at 66
represents power, its square root relates to an output voltage of
the surgical signal for a given impedance or resistance load. The
signal at 104 is thus a non-linear (square root) function of the
selected power signal at 66.
A scaling function is performed on the signal at 104 by a
conventional analog switch 106 and a resistor-divider network. The
scaling control signals from the mode logic circuit 62 (FIG. 1) are
supplied at 64 to selectively control a conventional analog switch
106 of the scaling circuit 60. The scaling control signals comprise
a plurality of individual signals, but for simplicity of
description each is referenced at 64. Upon application of a scaling
control signal at 64, one of the switches 106A or 106B is closed
and a voltage divider network is established between one of the
resistors 108 or 110 and the resistor 112. The one of the switches
106A or 106B which is closed depends on the mode of operation of
the electrosurgical unit selected by the surgeon. For simplicity of
description, only two different scaling functions are obtained from
the analog switch 106, although in reality a greater number will be
provided in accordance with at least the three different modes of
operation of the electrosurgical generator. The level of the limit
signal is established by the resistor-divider network.
The minimum current limit signal at 56, which limits the maximum
output voltage of the surgical signal, is supplied to the positive
input of a precision clamp 114 of the limit circuit 52. The RMS
current-related signal at 48 is supplied to the negative input of
the clamp 114. So long as the RMS current-related signal at 48
exceeds the minimum current limit signal at 56, the RMS
current-related signal at 48 is present at 68 as the current
delivered signal. However, should the RMS current-related signal at
48 fall below the minimum current limit signal at 56, the clamp 114
supplies the minimum current limit signal at 68 as the current
delivered signal. Thus, even though the electrosurgical generator
may actually be supplying less than the predetermined minimum
current in the surgical signal, the power regulation circuitry
operates on the artificial basis that the minimum current is
supplied. The maximum output voltage of the surgical signal is
limited accordingly. The effect is that the actual output power of
the electrosurgical generator rolls off or decreases into high
impedances because the power regulation feedback circuit operates
on the artificial basis of a constant output current delivery at
high impedances, due to the introduction of the minimum current
limit signal at 56 into the power calculation at the multiplier 72
instead of the RMS current-related signal at 48.
Examples of the actual power roll-off in the surgical signal from
the electrosurgical generator at high impedances, by using a
minimum current limit signal related to the square root of the
selected power signal or level, are shown by the curves 5A, 5B, 5C
and 5D in FIG. 5. The four curves 5A, 5B, 5C and 5D represent
selected power settings for the electrosurgical generator of one
hundred percent, seventy-five percent, fifty percent and
twenty-five percent, respectively. The curve 5A therefore
represents the maximum power output capability of the
electrosurgical generator. By deriving the minimum current limit
signal from the square root of the selected power signal, as has
been described and shown in FIG. 2, the roll-off in power
regulation capability at any selected power level occurs at
approximately the same predetermined relatively high impedance
designated ZH in FIG. 5 and occurs non-linearly generally like that
inherent non-linear power roll-off at maximum power delivery
capacity.
In many applications, it is desirable to avoid power roll-off at
high impedances when the electrosurgical generator is operating at
less than its maximum selected power capability. To avoid the
roll-off in power shown in FIG. 5 at the high impedances, when
operating the generator at less than its maximum output power
level, the current limit circuit 52 is eliminated and the minimum
current limit signal at 56 is not created. Instead, the RMS
current-related signal 48 is directly supplied as the current
delivered signal at 68 to the multiplier. Power regulation curves
6A, 6B, 6C and 6D illustrated in FIG. 6 result. Curve 6A represents
the inherent maximum power delivery capacity of the generator and
is essentially the same as curve 5A in FIG. 5. Curves 6B, 6C and 6D
represent the power output at seventy-five percent, fifty percent
and twenty-five percent of maximum capacity, respectively. At the
less-than-maximum capacity, constant or regulated power is
delivered into impedances greater than impedance ZH. Regulated
power is delivered until the maximum delivery capacity of the
generator is reached, i.e., when curves 6B, 6C or 6D intersect
curve 6A, at which point power roll-off occurs because the inherent
maximum power generation capacity is reached.
Attaining constant power regulation at high impedances at less than
maximum selected power output levels is an important improvement in
electrosurgery. It has been discovered that many beneficial effects
occur as a result of constant power regulation as the tissue
impedance increases or when relatively high impedance tissues are
encountered during the electrosurgical procedure. A better surgical
effect can be created by the surgeon as a result of this constant
power regulation. The pulse width modulation technique is more
effective for power regulation into the higher load impedances than
known prior power regulation techniques.
In some other situations, it is desirable to retain the limit
circuit 52 and generate a minimum current limit signal at 56, but
modify the value and relationship of the minimum current limit
signal to other signals and operative constraints of the generator.
For example, it may be desirable to limit the maximum output
voltage of the surgical signal to prevent or reduce flash and the
risk of alternate path burns but still obtain constant power
regulation into high impedance tissues. A circuit portion shown in
FIG. 7 is an example of a circuit which will create a constant
minimum limit signal at 56. With reference to FIG. 2, the op amp
100 and the square root network 102 are eliminated, and the circuit
portion shown in FIG. 7 is substituted. The signal at 104 is
directly connected to a constant positive circuit voltage. The
resistive network established by the resistors 108, 110 and 112,
and the selective closure of one of the switches 106A or 106B
establishes the minimum current limit signal at 56. An example of a
circuit which creates a limit signal which varies linearly with
respect to another variable signal is illustrated by the following
description of the voltage limit circuit 54, with the understanding
that the same principle can be applied in the creation of minimum
current limit signals.
Various types of minimum current limit signals at 56 have thus been
described. A minimum current limit signal which varies in
non-linear relationship (e.g., a square root relationship) to a
variable signal (e.g., the selected power signal at 66) is derived
from the circuit portion illustrated in FIG. 2. A constant minimum
current signal regardless of power setting is derived from the
circuit portion illustrated in FIG. 7. A linearly changing minimum
current limit signal is illustrated by the following description of
the derivation of the minimum voltage limit signal at 58. From
these examples, it is apparent that circuits for generating minimum
specially tailored current limit signals are possible. Such
circuits could regulate the power output capability at less than
maximum power settings to accommodate particular types of surgical
procedures, should it be discovered that particular types of
surgical procedures require specifically tailored power regulation
curves at particular impedances.
To obtain the desired output power signal at 76 as shown in FIG. 2,
the selected power signal at 66 is scaled as a result of an analog
switch 116 operatively controlled by the scaling control signals
applied at 64, in accordance with the selected mode of operation.
Closure of switch 116A causes the full selected power signal to be
applied to the op amp 118 which functions as a buffer. The desired
power output signal at 76 is the same as the selected power signal
at 66 under such circumstances. Closure of switch 116B establishes
a voltage divider network comprising resistors 120 and 122 to
reduce the magnitude of the selected power signal at 66 and cause
the desired power output signal at 76 to correspond to this reduced
level.
The minimum voltage limit signal at 58 is derived from the desired
output power signal at 76. The desired output power signal at 76 is
selectively switched into a voltage dividing network comprising
resistors 124, 126 and 128 by an analog switch 130 of the scaling
circuit 60. The switches 130A and 130B are selectively controlled
by the scaling control signals applied at 64. The minimum voltage
limit signal at 58, which operatively controls the maximum output
current of the surgical signal, is linearly related to the desired
output power signal at 76 due to the effects of the voltage divider
network.
An op amp 132 functions as a precision clamp in the limit circuit
54. The minimum voltage limit signal at 58 is applied to the
positive terminal of the op amp 132 and the RMS voltage-related
signal at 50 is applied to the negative terminal. So long as the
RMS voltage-related signal at 50 is greater than the minimum
voltage limit signal at 58, the RMS voltage-related signal is
supplied as the voltage delivered signal at 70. However, should the
RMS voltage-related signal at 50 fall below the minimum voltage
limit signal at 58, the minimum voltage limit signal is supplied as
the voltage delivered signal at 70.
By introducing the minimum voltage limit signal as an artificial
substitute for the RMS voltage-related signal, the maximum output
current of the surgical signal is limited to a maximum value even
though the output impedance may actually be so low at a much larger
output current should actually flow from the electrosurgical
generator. For any desired output power level, a minimum voltage
level signal is established which linearly relates to that desired
output power level. Because the minimum voltage limit signal at 58
establishes that constant maximum output current of the surgical
signal which the electrosurgical generator will deliver into low
impedances the minimum voltage limit signal and the desired output
power signal at 76 are linearly related. The output current will be
limited to a predetermined maximum at all low impedances,
regardless of power settings. This can be understood by reference
to the low impedance ranges of the graphs of FIGS. 5 and 6. The
output surgical power increases approximately linearly as the
impedance increases in the low impedance range (up to ZL) because
of the constant maximum value which the current can attain at low
impedances due to the introduction of the artificial minimum
voltage limit signal at 58 related to the desired power output
level. The limit on the maximum output current prevents internal
destruction of circuit elements of the generator, among other
advantages.
The current and voltage delivered signals at 68 and 70,
respectively, are applied to the input terminals of a conventional
multiplier 72 as shown in FIG. 2. These signals are multiplied
together and the product signal is supplied as a delivered power
signal at 74 to the positive input terminal of differential
amplifier 78. The desired output power signal at 76 is applied
through an appropriate resistance network to the negative input
terminal of the differential amplifier 78. The differential
amplifier 78 supplies an error signal at 80 which is related in
magnitude and sign (positive or negative) to the difference between
the delivered power signal at 74 and the desired output power
signal at 76. When there is a great disparity between the delivered
and desired amounts of power, the magnitude of the error signal at
80 is great. When the delivered power is approximately equal to the
desired power, the magnitude of the error signal at 80 is very
small or substantially nonexistent. The sign of the error signal at
80 establishes whether more or less power should be supplied to
achieve regulation.
The RMS to DC converters 44 and 46 are conventional items, as is
the multiplier 72. RMS to DC converters which have proved
satisfactory are number AD 536 AJH, manufactured by Analog Devices
of Two Technology Way, P.O. Box 280, Norwood, Mass., 02062,
U.S.A.
Details of the pulse width modulation circuit 82, the amplifier
drive circuit 86, the RF amplifier 82 and the output transformer 26
are shown in FIG. 3. The error signal at 80 from the differential
amplifier 78 (FIGS. 1 and 2) is applied to a conventional
integrator defined by an op amp 134 and an integrating feedback
network including the capacitor 136. The integrator has the effect
of continually time integrating or averaging the error signal 80,
as well as creating control loop stability. The output signal of
the integrator at 138 is always a positive level trigger level
signal. The sign of the error signal created by the differential
amplifier 78 (FIG. 2) is coordinated with the operation of the
integrator to create this positive level trigger signal. When the
error signal at 80 is negative in sign, indicating a need for more
power, the integration increases the magnitude of the trigger level
signal at 138. When the error signal at 80 is positive in sign,
indicating the need for less power, the integration decreases the
magnitude of the trigger level signal at 138. When the error signal
at 80 is zero or nonexistent, the magnitude of the trigger level
signal at 138 remains unchanged.
The trigger level signal at 138 is presented to the base terminal
of a transistor 140. Transistor 142 and transistor 140 form a
discrete component comparator. The other input signal to this
discrete comparator is applied at 144 to the base terminal of the
transistor 142. This other input signal at 144 is that signal
across capacitor 146. The transistor 148 and its associated biasing
elements define a constant current source for charging the
capacitor 146 at a constant current rate. Accordingly, the voltage
signal across capacitor 146 increases in a linear or ramp-like
fashion and thus creates a ramp signal at 144. A signal at 150 from
a conventional edge detector 152 energizes the FET 154 to discharge
the capacitor 146. Once discharged, the capacitor 146 immediately
commences charging again.
The ramp signal at 144 across the capacitor 146 is periodic in
nature, because the edge signal at 150 is periodic, and the
capacitor 146 periodically discharges through the FET 154. The
periodic edge signal at 150 is derived from the oscillator signal
at 98 supplied from a conventional oscillator 156 which is a part
of the amplifier drive circuit 86. The oscillator signal at 98
establishes the frequency for the high or radio frequency surgical
signal delivered to the patient by the electrosurgical generator.
The oscillator signal at 98 is shown in FIG. 4A. The edge detector
152 responds to each positive going and negative going edge of the
oscillator signal and supplies a narrow pulse at each edge
transition of the oscillator signal. The edge signal shown in FIG.
4D is thus a series of relatively narrow pulses, each occurring at
an edge of the oscillator signal. Each pulse of the edge signal
causes the FET 154 to rapidly discharge the capacitor 146. The
constant current source established by the transistor 148
immediately commences charging the capacitor 146 and the voltage
across the capacitor builds linearly to create the ramp signal at
144 shown in FIG. 4E. Thus, the ramp signal shown in FIG. 4E takes
on the characteristics of a sawtooth wave having a frequency
established by the edge signal and which is approximately twice the
frequency of the oscillator signal shown in FIG. 4A.
The oscillator signal at 98 is presented to a flip-flop logic and
gating circuit 160 and to the duty cycle generator 92, as shown in
FIG. 3. The duty cycle generator 92 is under the control of the
mode logic circuit 62 (FIG. 1) by virtue of the signals at 94, and
establishes the duty cycle signal at 96 to control the delivery of
the high-frequency pulses in accordance with the selected mode of
operation. The duty cycle signal at 96 is referenced to and
coordinated with the oscillator signal at 98 to cause the on-time
and off-time periods of the duty cycle envelope to begin with and
end with the oscillator cycles. So long as the duty cycle generator
92 signals at 96 for the delivery of the high-frequency surgical
signal, the logic and gating circuit supplies two periodic pulse
phase signals at 162 and at 164 at the predetermined high or radio
frequency of the oscillator signal at 98. The two pulse phase
signals are phase-shifted one hundred eighty degrees with respect
to one another. A pulse phase 1 signal is present at 162 and a
pulse phase 2 signal is present at 164. The width of each pulse in
both the pulse phase 1 and phase 2 signals represents the maximum
width to which each driving pulse at 90 (FIGS. 1 and 3) is allowed
to expand to achieve power regulation. The pulse phase 1 signal and
the pulse phase 2 signal are represented at FIGS. 4B and 4C,
respectively.
The technique for achieving pulse width modulation by virtue of the
trigger level signal at 138 can now be described. Initially, the
edge signal at 150 causes the FET 154 to discharge the capacitor
146. Thereafter, the capacitor 146 commences charging and
transistor 142 begins conducting. Transistor 142 continues to
conduct as the voltage across capacitor 146 reaches a level
equivalent to the level of the trigger level signal at 138. As soon
as the voltage across capacitor 146, i.e., the ramp signal at 144,
increases slightly over the trigger level signal, transistor 140
commences conducting and transistor 142 stops conducting, because
the voltage on the base terminal of transistor 142 has exceeded the
voltage at the base terminal of transistor 140. Once transistor 140
commences conducting a termination signal is present at 166 across
resistor 168 and at the base of transistor 170. The termination
signal at 166 is illustrated in FIG. 4G.
The effects of the trigger level signal at 138 in controlling the
ramp signal at 144 due to the action of the discrete component
comparator formed by transistors 140 and 142, is illustrated in
FIG. 4F. As soon as the ramp signal increases to a level equivalent
to the trigger level signal, the termination signal shown in FIG.
4G is delivered. The width of each pulse of the termination signal
is that remaining time portion of each interval of the ramp signal
(FIG. 4E) before discharge the capacitor 146 and the commencement
of the next individual ramp of the ramp signal. The high portion of
the termination signal at 166 biases the transistor 170 into
conduction.
The pulse width control signal at 84 is created by the switching
effects of transistor 170. The signal level at 84 immediately drops
when transistor 170 begins conducting due to the effects of the
resistor 174. When transistor 170 is not conducting, the level of
the signal at 84 is high. The pulse width control signal is
illustrated in FIG. 4H. The pulse width control signal is the
inversion of the termination signal shown in FIG. 4G.
The pulse width control signal at 172 is applied to one input
terminal of both AND gates 176 and 178. The pulse phase 1 signal at
162 is applied to the other input terminal of the AND gate 176 and
the pulse phase 2 signal at 164 is applied to the other input
terminal of another AND gate 178. AND gates 176 and 178 supply high
output signals at 180 and 182, respectively, so long as both input
signals are high. A pulse width modulated phase 1 signal is present
at 180 upon the existence of the high level of the pulse phase 1
signal at 162 and the existence of high level of the pulse width
control signal at 84. The pulse width modulated phase 1 signal at
180 goes to a low level when the pulse width control signal at 84
drops to a low level. Accordingly, the time width of the pulse
width modulated phase 1 signal is controlled or modulated by the
pulse width control at 84. This is illustrated by considering that
the signals shown in FIGS. 4B and 4H are both at high levels during
the time that the pulse width modulated phase 1 signal shown in
FIG. 4I is delivered. As soon as the pulse width control signal
shown in FIG. 4H goes low, the pulse width modulated phase 1 signal
also goes low. A similar situation exists with respect to the pulse
width modulated phase 2 signal at 182. The AND gate 178 gates the
pulse phase 2 signal at 164 (FIG. 4C) with the pulse width control
signal at 84 (FIG. 4H). The width of each pulse width modulated
phase 2 signal at 182 terminates when the pulse width control
signal goes low. The pulse width modulated phase 2 signal is shown
in FIG. 4J and is derived by considering FIGS. 4C and 4H in the
logical manner established by operation of the AND gate 178.
It should be noted that the edge signal at 150 controls the FET 184
simultaneously with the FET 154. When the FET 184 is conductive,
the signal level at 166 drops approximately to reference level and
the conduction of transistor 170 terminates. Thus, conduction of
the FET 184 assures that the pulse width control signal at 84
commences each pulse width determination period at a high level and
also assures that transistor 142 is conducting at the beginning of
each pulse width determination period.
As has been described, the error signal at 80 and the trigger level
signal at 138 operatively control the width of each pulse width
modulated phase 1 and phase 2 signal at 180 and 182, respectively.
When the error signal at 84 is substantially large in a negative
sense, indicating the need for great power, the ramp signal (FIG.
4E) present at 144 will not reach the relatively large magnitude of
the trigger level signal, in contrast to that situation shown in
FIG. 4F. Hence, substantially full width pulse width modulated
phase 1 and phase 2 signals will be delivered at 180 and 182
because the transistor 140 will not become conductive. The edge
signal at 150 will cause capacitor 146 to discharge before
transistor 140 ever becomes conductive. Since transistor 140 never
becomes conductive, the pulse width control signal at 84 remains
continually high and the width of each pulse of the pulse width
modulated phase 1 and phase 2 signals at 180 and 182, respectively,
is driven to the full width of the pulse phase 1 and pulse phase 2
signals at 162 and 164, respectively. Accordingly, FIGS. 4B and 4C
also respectively represent the full width pulse width modulated
phase 1 and phase 2 signals present both at 180 and 182. As soon as
power builds up and the error signal 80 decreases to zero, the
level of the trigger level signal attains desired power regulation
because the width of the pulses is established to secure the
desired amount of power delivery. If the electrosurgical generator
is delivering an excessive amount of power, the error signal at 80
becomes positive. The integration of the positive error signal
reduces the magnitude or level of the trigger level signal at 138,
thus causing the pulse width control signal (FIG. 4H) to drop to a
low level at an earlier point in each full phase time period.
Accordingly, the width of each pulse width modulated phase 1 and
phase 2 signal is reduced and the amount of output power is thus
reduced.
In addition to those functions of the flip-flop logic and gating
circuit 160 previously described, the flip-flop logic and gating
circuit also includes conventional gating circuit elements (not
shown) for assuring that pulse width modulated phase 1 signal at
180 is delivered first, followed by a pulse width modulated phase 2
signal at 182. In addition, when the duty cycle generator 92 calls
for the termination of the surgical signal, the logic and gating
circuit 160 assures that the on time of the duty cycle envelope
terminates after a pulse width modulated phase 2 signal has been
delivered. All of the functions of the flip-flop logic and gating
circuit 160 can be achieved by the interconnection of binary logic
elements, primarily flip-flops and gates.
Each of the pulse width modulated phase 1 and phase 2 signals at
180 and 182, respectively, is applied to its own phase drive
circuit. One phase drive circuit is illustrated at 186. The phase
drive circuits for both the pulse width modulated phase 1 and phase
2 signals are the same as that single one illustrated at 186.
Accordingly, a description of the operation of the phase drive
circuit 186 is made below with respect to a pulse width modulated
phase signal P, although it should be understood that both the
pulse width modulated phase 1 and phase 2 signals have the same
effect on their respective phase drive circuits as the phase signal
P has on the phase drive circuit 186.
The phase signal P is applied at l8S to the phase drive circuit 186
and causes FET 190 to become conductive. A transformer 192 includes
a center tapped primary winding and the coil 194 thereof is poled
to induce a positive signal at terminal 196 with respect to
terminal 198 and a positive signal at terminal 200 with respect to
terminal 202. The terminals 196 and 200 are connected to FET's Q1A
and Q1B of the RF amplifier 22. The positive signals at 196 and 200
turn on both FET's Q1A and Q1B and causes current to be conducted
at 20 from the DC power supply 16 (FIG. 1) through the primary
winding of the power output transformer 26. Whenever the phase
signal P terminates, a narrow reset pulse P goes high at conductor
204. The reset pulse signal P is created by the negative going edge
of the phase signal P. FET 206 becomes conductive and current is
momentarily conducted in the reverse direction in the primary
winding coil 208 of the transformer 192. The narrow reverse pulse
of current in the primary winding coil 208 resets the magnetics or
hysteresis characteristics of the core of the transformer 192 to
ready it for conduction during the next phase signal P. The various
signals at terminals 196, 198, 200 and 202 are illustrative of
those comprising, collectively, the drive signal at 90.
The other one of the two pulse width modulated phase signals at 180
or 182 has a corresponding effect on its phase drive circuit and
the FET's Q2A and Q2B are rendered conductive and nonconductive in
the same manner as has been previously described. When FET Q2A and
Q2B are conductive, the direction of current flow through the
primary winding of the power output transformer 26 reverses.
Accordingly, an alternating current pulse width modulated signal at
28 is created by the drive signal applied to the amplifier 22.
Examples of the alternating pulse width modulated signal at 24 are
shown in FIGS. 4K and 4L.
The alternating pulse width modulated signal to the primary winding
of the power output transformer 26 for full-width driving pulses of
the drive signal at 90 is illustrated in FIG. 4K. In the waveform
shown in FIG. 4K, it is to be noted that the full-width pulse width
modulated phase 1 signal (e.g., FIG. 4B) creates the positive
portion of the signal and the full-width pulse width modulated
phase 2 signal (e.g., FIG. 4C) creates the negative portion of the
transformer input signal. For less-than-full-width driving pulses
of the drive signal, the waveform presented to the primary winding
of the power output transformer is shown in FIG. 4L. Again, the
pulse width modulated phase 1 signal (FIG. 4I) creates the positive
portion while the pulse width modulated phase 2 signal (FIG. 4J)
creates the negative portion. It is to be noted that the waveform
shown in FIG. 4L has the frequency characteristic exactly the same
as the frequency characteristic of the oscillator signal (FIG.
4A).
The amount of energy delivered by the AC pulse modulated signal at
28 from the power transformer 26 is defined generally by the area
above and below the zero reference point of the waveforms shown in
FIGS. 4K and 4L, although the AC pulse modulated signal at 28 will
not actually have the square pulse shapes shown due to the
inductive effects of the filter 30 which are reflected back to the
primary winding of the transformer 26. This energy is presented at
a periodic basis at the band pass frequency of the band pass filter
30 (FIG. 1). Accordingly, the band pass filter is driven at its
band pass frequency to deliver the sinusoidal surgical signal shown
in FIG. 4M at the predetermined high frequency. The passive
reactive elements of the band pass filter 30 change the AC pulse
modulated signal at 28 into sinusoidal oscillations. Each cycle of
the sinusoidal surgical signal is created by and correspondingly
results from one cycle (a positive and negative pulse) of the pulse
width modulated signal at 24, e.g., FIGS. 4K and 4L. The
relationship and correspondence between the pulse width modulated
signal at 24 and the sinusoidal surgical signal at 32 is
illustrated by comparing FIG. 4M to FIGS. 4K and 4L. When a
full-width pulse-width-modulated signal is received, such as that
shown in FIG. 4K, the amplitude of the sinusoidal surgical signal
will be greater than when a less-than-full-width
pulse-width-modulated signal, such as that shown in FIG. 4L, is
supplied for the same impedance load. Thus, the power of the
surgical signal present on conductor 32 is defined by the area or
width of the pulse width modulated phase 1 and phase 2 signals and
the corresponding pulses of the drive signal which drive and
control the switching of the amplifier 22.
One of the advantages of regulating both the output of the DC power
supply 16 by the control signal at 18, shown in FIG. 1, and by
pulse width modulation as described herein is that the pulse width
modulation obtains a better resolution (i.e., allows expansion to
substantially the major portion of the pulse width) for given power
settings. In other words, the DC power supply 16 (FIG. 1) generally
or coarsely regulates the amount of power and the pulse width
modulation capability of the present invention achieves a finally
regulated and rapid control over the amount of power actually
delivered. The inherent maximum power delivery capacity of the
power supply is limited by this approach, however, and relatively
rapid power roll-off occurs at higher output impedances.
The pulse width modulation power regulation technique described
herein allows the energy content of each cycle of the sinusoidal
output wave applied to the patient to be energy regulated. Very
precise power regulation occurs. Very rapid response times are also
possible to achieve greatly improved constant power regulation when
the tissue impedance rapidly fluctuates. Superior and greatly
improved surgical effects result. The constant power regulation
available from the present invention even into relatively high
impedance tissues is a substantial improvement in the field of
electrosurgery. Limiting the maximum output voltage at high
impedances in the manner described herein avoids or reduces the
possibility for flash and undesirable arcing, as well as reducing
the risk of alternate path burns to the patient. Limiting the
maximum output current at low impedances to a predetermined maximum
at any particular power setting avoids the possibility of
destruction to the electrosurgical generator as a result of short
circuiting the output electrodes or terminals. Numerous other
improvements and advantages of the present invention have been
discussed above or will be apparent after full comprehension of the
present invention.
A preferred embodiment of the present invention has been shown and
described with a degree of particularity. It should be understood,
however, that the specificity of the present description has been
made by way of preferred example, and that the scope of the present
invention is defined by the appended claims.
* * * * *